The fact that (i) 80% of hepatocellular carcinomas (HCCs) arise in fibrotic livers and that (ii) chronic injury, inflammation and fibrosis are sufficient to induce HCC suggest that HCC truly represents "a wound that does not heal". Here we seek to investigate how myofibroblast activation promotes HCC with the ultimate goal to define "druggable" targets in the hepatic tumor microenvironment (TME). In the first two Aims we will focus on upstream pathways that activate a tumor-promoting phenotype in myofibroblasts. We have previously demonstrated a crucial role for the gut microbiota and TLR4 in hepatic myofibroblast activation. Our preliminary data show promotion of HCC by the gut microbiota and TLR4. We hypothesize that the tumorpromoting effects of the gut microbiota and TLR4 are mediated by myofibroblasts and will test this hypothesis by investigating hepatocarcinogenesis and myofibroblast activation in mice with myofibroblastspecific TLR4 deletion and in germ-free mice (Aim 1). We hypothesize that chronic epithelial injury promotes HCC through the release of damage-associated molecular patterns (DAMPs) and DAMP-mediated myofibroblast activation. We will test the hypothesis that the DAMPs ATP and HMGB1 promote myofibroblast activation and HCC in mice with deletion of the ATP receptor P2X7 or a hepatocyte-specific HMGB1 deletion (Aim 2). In last two Aims, we will focus on effector pathways through which myofibroblasts promote HCC. To determine whether myofibroblasts promote HCC through extracellular matrix (ECM), we will investigate whether spontaneous liver fibrosis increases HCC using a model in which Lhx deletion increases ECM synthesis and cross-linking (Aim 3). To determine whether inflammatory signals in myofibroblasts promote HCC, we will assess hepatocarcinogenesis in mice with deletion or activation of IKK(3 in myofibroblasts and in mice that lack epiregulin, an NF-KB-regulated hepatomitogen with high expression in hepatic myofibroblasts (Aim 4). Detailed studies in human myofibroblasts and human HCC will confirm the clinical relevance of all above described pathways. We anticipate that the proposed studies will reveal druggable targets in the TME and thus lead to novel concepts for the treatment or prevention of HCC.